Portail HAL de l'Université du Littoral Côte d'Opale
Not a member yet
11255 research outputs found
Sort by
The bi-dimensional directed IDLA forest
No full textPublié dans Annals of Applied Probability, Vol. 33, No. 3, p. 2247-2290, 2023We investigate three types of Internal Diffusion Limited Aggregation (IDLA) models. These models are based on simple random walks on Z 2 with infinitely many sources that are the points of the vertical axis I (∞) = {0} × Z. Various properties are provided, such as stationarity, mixing, stabilization and shape theorems. Our results allow us to define a new directed (w.r.t. the horizontal direction) random forest spanning Z 2 , based on an IDLA protocol, which is invariant in distribution w.r.t. vertical translations
Step-by-step framework for automated ageing of plaice using otolith shape
No full textInternational audienceOtoliths are reliable biological markers used to determine fish age and other ecological characteristics. Their shape and structure are influenced by genetic, environmental, ontogenic, and life-stage factors, all of which affect otolith formation and growth. This study aims to develop an automated approach for fish ageing based on quantitative analysis of external otolith shape. By extracting shape descriptors from otolith images, the method enables age estimation without the need for manual counting of growth rings a traditionally labor-intensive process requiring expert interpretation.A systematic framework was established for preprocessing and standardizing otolith images prior to age estimation. Images are binarized, contours are extracted, and orientation is standardized through rotation correction, side detection, and mirror transformation. Multivariate shape features are then derived using advanced mathematical representations, including elliptic Fourier descriptors, wavelet transforms, and curvature scale space methods. Univariate parameters such as length, width, height, and area are also calculated, both for the entire contour and for each of the four otolith sides, enabling localized shape analysis. The extracted shape features serve as input for several supervised machine learning models Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Recursive Partitioning (RPART), and linear models to predict fish age. The dataset comprises thousands of otolith images representing age classes from 0 to 7 years and older. Model performance is evaluated using Root Mean Square Error (RMSE) to quantify prediction accuracy.Results show that the KNN model (k = 7) achieved the best performance, with approximately 52% accuracy within ±1 year of the true age. While the findings are promising, some variability persists, particularly in distinguishing between middle and older age classes, where ontogenic effects on otolith shape become less pronounced
Complementary aerosol mass spectrometry elucidates sources of wintertime sub-micron particle pollution in Fairbanks, Alaska, during ALPACA 2022
No full textInternational audienceFairbanks, Alaska, is a subarctic city that frequently suffers from non-attainment of national air quality standards in the wintertime due to the coincidence of weak atmospheric dispersion and increased local emissions. However, significant uncertainties exist about aerosol sources, formation, and chemical processes during cold winter periods. We aim to determine the composition, size, and concentrations of atmospheric sub-micron non-refractory particulate matter (NR-PM1) and quantify their sources in the urban centre of Fairbanks. As part of the Alaskan Layered Pollution and Chemical Analysis (ALPACA) campaign, we deployed a Chemical Analysis of Aerosol Online (CHARON) inlet coupled with a proton transfer reaction – time of flight mass spectrometer (PTR-ToF MS) and an Aerodyne high-resolution aerosol mass spectrometer (AMS) to measure organic aerosol (OA) and NR-PM1, respectively. We used positive matrix factorisation (PMF) for source identification. PTRCHARON factorisation delineated four residential heating sources, including wood and oil combustion, that contribute 47 ± 20 % and 16 ± 9 % of OACHARON, on average, respectively. In contrast, only a single biomass burning-related factor was identified by AMS for both OA and NR-PM1, but it provided information on two additional factors that were rich in sulphur and nitrate. These results demonstrate that PTRCHARON can generate robust quantitative information with enhanced resolution of organic aerosol sources. When combined with suitable complementary instruments like the AMS, such evidence-based insights into the sources of sub-micron aerosol pollution can assist environmental regulators and citizen efforts for the improvement in air quality in Fairbanks and in the wider Arctic winter
SBA-15 supported Ni-Cu catalysts for hydrodeoxygenation of m-cresol to toluene
No full textInternational audienceAmidst concerns over fossil fuel dependency and environmental sustainability, the utilization of biomass-derived aromatic compounds emerges as a viable solution across diverse industries. In this scheme, the conversion of biomass involves pyrolysis, followed by a hydrodeoxygenation (HDO) step to reduce the oxygen content of pyrolysis oils and stabilize the end products including aromatics. In this study, we explored the properties of size controlled NiCu bimetallic catalysts supported on ordered mesoporous silica (SBA-15) for the catalytic gas-phase HDO of m-cresol, a lignin model compound. We compared their performances with monometallic Ni and Cu catalysts. The prepared catalysts contained varying Ni to Cu ratios and featured an average particle size of approximately 2 nm. The catalytic tests revealed that the introduction of Cu alongside Ni enhanced the selectivity for the direct deoxygenation (DDO) pathway, yielding toluene as the primary product. Optimal performance was observed with a catalyst composition comprising 5 wt.% Ni and 5 wr.% Cu, achieving 85 % selectivity to toluene. Further increasing the Cu content improved turnover frequency (TOF) values, but reduced DDO selectivity. These findings underscore the importance of catalyst design in facilitating biomass-derived compound transformations and offer insights into optimizing catalyst composition for more selective HDO reactions
identifications des peptides à courtes chaînes en agroalimentaire : challenges et aperçus critiques
No full textInternational audienceAs the global population continues to grow, optimizing the use of the protein resources produced by our planet has become essential for supporting the nutrition and health of both humans and animals. Over the years, the agro-food industry has played a key role in valorizing protein-rich by-products generated during food processing. Fermentation and enzymatic hydrolysis of these by-products are efficient, industry-friendly, and cost-effective methods for recovering peptides in the form of peptide hydrolysates. These hydrolysates contain a range of protein compounds, including large-, medium-, and short-chain peptides (SCPs), as well as amino acids, many of which are bioactive. Numerous databases now catalog food-derived bioactive peptides, with the BioPep-UWM database being one of the most comprehensive, containing 5,014 bioactive peptides (as of July 2024) associated with 90 distinct bioactivities. Of these, nearly 45% are SCPs, including 581 di-, 754 tri-, 385 tetra-, and 511 penta-peptides, totaling 2,231 SCPs. Despite their prevalence, the identification of SCPs remains more challenging than that of large- or medium-chain peptides, and research dedicated to SCP identification began only about twelve years ago. This review critically examines the technical challenges involved in SCP identification, offering a detailed timeline of approximately fifty key studies that have contributed to advancing the field. Specifically, the review explores enrichment and clean-up protocols, analytical methods, mass spectrometry techniques, and the bioinformatics approaches used for tandem mass spectrometry data processing in the context of SCP identification.Alors que la population mondiale continue de croître, l'optimisation de l'utilisation des ressources protéiques produites par notre planète est devenue essentielle pour soutenir la nutrition et la santé des humains et des animaux. Au fil des ans, l'industrie agroalimentaire a joué un rôle clé dans la valorisation des sous-produits riches en protéines générés lors de la transformation des aliments. La fermentation et l'hydrolyse enzymatique de ces sous-produits sont des méthodes efficaces, respectueuses de l'industrie et rentables pour récupérer les peptides sous forme d'hydrolysats de peptides. Ces hydrolysats contiennent une gamme de composés protéiques, y compris des peptides à chaîne large, moyenne et courte (SCP), ainsi que des acides aminés, dont beaucoup sont bioactifs. De nombreuses bases de données répertorient désormais les peptides bioactifs d'origine alimentaire, la base de données BioPep-UWM étant l'une des plus complètes, contenant 5 014 peptides bioactifs (en juillet 2024) associés à 90 bioactivités distinctes. Parmi ceux-ci, près de 45 % sont des SCP, dont 581 di-, 754 tri-, 385 tétra- et 511 penta-peptides, soit un total de 2 231 SCP. Malgré leur prévalence, l'identification des SCP reste plus difficile que celle des peptides à chaîne large ou moyenne, et la recherche consacrée à l'identification des SCP n'a commencé qu'il y a une douzaine d'années. Cette étude examine de manière critique les défis techniques liés à l'identification des SCP et propose une chronologie détaillée d'une cinquantaine d'études clés qui ont contribué à faire avancer le domaine. Plus précisément, l'étude explore les protocoles d'enrichissement et de nettoyage, les méthodes analytiques, les techniques de spectrométrie de masse et les approches bioinformatiques utilisées pour le traitement des données de spectrométrie de masse en tandem dans le contexte de l'identification des SCP
Une étude spectroscopique operando de l’adduit Re2O7.2H2O obtenu par hydratation contrôlée de Re2O7.
No full textInternational audienceWe provide here a comprehensive investigation spectroscopic of the controlled hydration of Re2O7 using Raman, FTIR and XAS techniques in complement with ab initio modelling for confirming the spectral assignments. Hence, the Raman signature of Re2O7.2H2O was obtained, and the evolution kinetics was investigated to provide a detailed description of the hydration process
Introducing the "Cockpit Party Problem": Blind Source Separation Enhances Aircraft Cockpit Speech Transcription
No full textInternational audienceCockpit Voice Recorders (CVRs) are one of the two mandatory flight recording devices embarked in commercial aircraft. Its analysis is crucial to understand the context of an air incident or accident. However, in such scenarios, when the audio recordings are usable, CVR may contain strong mixtures of crew member speech signals, radio communications, and cockpit alarms. However, contrary to the "cocktail party problem" that Blind Source Separation (BSS) aims to tackle, modeling CVR mixtures—that we here name the "cockpit party problem"—was never done before. In this paper, we thus propose a CVR mixture model and highlights its limitations. While not trivial—even in a two-source scenario—BSS methods can be applied to real CVR recordings. We find that taking into account several BSS outputs provided by various methods may help audio analysts to transcribe the CVR data. That is near 90% of unintelligible words can be transcribed from CVR recordings processed by BSS methods
Evaluation of CALIOP derived backscattering coefficient (<i>b</i><sub>bp</sub>(532)) product over contrasted oceanic waters
No full textInternational audienceCALIOP satellite sensor offers advantages over passive sensors, particularly during nighttime and in polar-subpolar regions. Though originally designed for atmospheric studies, it was adapted to retrieve the backscattering coefficient at 532 nm, bbp(532) in the ocean. Scarce matchups and the lack of standardized protocol hindered previous validation efforts. An evaluation using a standardized protocol and diverse in-situ datasets from contrasted oceanic waters was carried out for the period 2008-2021, with the 2018-2021 period being evaluated for the first time. A strong correlation was observed with R2 up to 0.94 (RMS: 0.001-0.01 m-1, MRE: 42.7%-63%, bias: 36.88%-13.09%). A comparison with MODIS-Aqua bbp(532) product was performed showing comparable estimates of 2. BGC-Argo data from 2018-2021 were also used to evaluate CALIOP bbp(532), revealing a lower correlation compared to 2008-2021 period, probably due to a decrease in CALIOP lidar power
L'essentiel des principes fondamentaux de droit constitutionnel
No full textNational audienceL'ouvrage aborde les fondements du droit constitutionnel (notions d’État, de souveraineté, de Constitution et de séparation des pouvoirs), les régimes constitutionnels étrangers, (États-Unis, Royaume-Uni et l’Allemagne). Ces notions théoriques sont mises en perspective par une analyse de l’histoire constitutionnelle française (évolution des concepts et notions du droit constitutionnel).Cet ouvrage, spécialement conçu pour les étudiants de licence, offre une approche complète et pédagogique permettant de maîtriser les fondamentaux et les subtilités de cette discipline essentielle. Chaque chapitre est rédigé de manière claire et structurée, avec des tableaux récapitulatifs et des exemples concrets.Quelques nouveautés en 2025Intègre les débats marquants la vie politique française comme la dissolution de juillet 2024, la motion de censure, le report de l'adoption de la loi de finances, ou encore les élections américaines, britanniques, allemandes...Conséquences de l’absence de majorité absolue à l’Assemblée nationale sur la pratique de la Ve RépubliqueLes décisions du Conseil constitutionnel, notamment sur la loi réformant les retraites et le référendum d’initiative partagé